3 ms·I’m sure someone is working on adding support for human-like errors in LLM outputs.by bleuchase 4y agoI’m sure someone is working on adding support for human-like errors in LLM outputs.